Design of Experiments (DOE) is a method for organizing and analyzing multivariable experiments in an efficient and rigorous manner. Applications of DOE are not restricted to physical tests, however. Results from computer “experiments” can also be analyzed to determine the sensitivity and significance of changes in input variables on the output (response) of any computerized process model. This work includes a brief coverage of some of the fundamentals of DOE, its terminology and some shortcuts in its application.

Examples from the field of accident reconstruction are presented that contain a large number of input variables. Under rather broad conditions and meeting proper assumptions, Planar Impact Mechanics (PIM) can be used to analyze collisions between two vehicles. Given each vehicle’s physical properties, the collision configuration and information about the contact process (coefficient of restitution and tangential impact coefficient), final impact velocities, collision kinetic energy loss, ΔV values and other results can be calculated algebraically when the initial velocities are specified. DOE is applied to the planar impact collision model for two collision configurations as a method to examine the uncertainty of energy loss and ΔV of typical collisions relative to the input variables.